^{1,a)}, Nobumasa Miyawaki

^{1}, Susumu Okumura

^{1}, Ikuo Ishibori

^{1}, Takayuki Nara

^{1}, Takashi Agematsu

^{1}, Ken-ichi Yoshida

^{1}, Watalu Yokota

^{1}, Yoshiteru Nakamura

^{1}, Kazuo Arakawa

^{1}and Mitsuhiro Fukuda

^{2}

### Abstract

Single-turn extraction from the Japan Atomic Energy Agency AVF cyclotron with a number of 110 using a flat-top (FT) acceleration system has been achieved to reduce the energy spread of an ion beam for microbeam formation with energy up to hundreds of MeV and to increase extraction efficiency from the cyclotron. In order to generate a FT waveform voltage using the fifth-harmonic frequency on a dee electrode, a FT resonator was designed using MAFIA code to achieve downsizing and low power consumption. The FT resonator, coupled to the main resonator through a coupling capacitor, covered the full range of the fifth harmonic frequency from 55 to 110 MHz. Various ion beams, accelerated using different acceleration harmonic modes of and 2, such as 220 MeV , 260 MeV , and 45 MeV , were developed by FT acceleration. A clear turn separation of the beam bunches was successfully observed at the extraction region of the large-scale AVF cyclotron with number of revolutions greater than 200. As a result, high extraction efficiency (over 95%) from the cyclotron was achieved. Single-turn extraction was confirmed by counting the number of beam bunches out of the cyclotron for an injected beam pulsed by a beam chopping system in the injection line. The energy spread of the 260 MeV beam was measured using an analyzing magnet, and we verified a reduction in the energy spread from to 0.05% by single-turn extraction after FT acceleration.

The authors thank Dr. Goto and Mr. Kohara at RIKEN for their kindness of lending the flat-top resonator, and Prof. Shinozuka at Tohoku University for his kind support of the preliminary low-level power test. The authors are also grateful to the operators of the cyclotron for assistance in tuning the beam and running the experiments.

I. INTRODUCTION

II. DESIGN OF THE FIFTH-HARMONIC RESONATOR

A. Design of the resonator using the MAFIA code

B. Performance of the FT resonator

III. PREPARATION OF THE CYCLOTRON FOR FT ACCELERATION

A. Tuning of the beam phase width and the magnetic field

B. Estimation of the fifth-harmonic voltage

IV. MEASUREMENT AND ANALYSIS OF FT-ACCELERATED BEAM

A. Measurement of turn separation

B. Single-turn extraction

C. Measurement of the energy spread

### Key Topics

- Electrodes
- 24.0
- Cyclotron resonances
- 18.0
- Ion beams
- 12.0
- Magnetic fields
- 9.0
- Magnets
- 8.0

## Figures

Schematic layout of main components of the JAEA AVF cyclotron, viewed by excluding the upper side of the main magnet, with a picture of the FT resonator. The FT resonator is capacitively coupled to the fundamental resonator.

Schematic layout of main components of the JAEA AVF cyclotron, viewed by excluding the upper side of the main magnet, with a picture of the FT resonator. The FT resonator is capacitively coupled to the fundamental resonator.

Schematic diagram of the resonators. The span angle of the dee electrode and its maximum voltage are 86° and 60 kV, respectively. A fundamental resonance at frequencies ranging from 11 to 22 MHz is obtained by adjusting the position of the movable short L1 within a range of 1350 mm.

Schematic diagram of the resonators. The span angle of the dee electrode and its maximum voltage are 86° and 60 kV, respectively. A fundamental resonance at frequencies ranging from 11 to 22 MHz is obtained by adjusting the position of the movable short L1 within a range of 1350 mm.

Dependence of the fundamental and resonance frequencies on the position of the movable short L1. Open circles and closed triangles represent measured and calculated frequencies, respectively. The fundamental mode at is used as the acceleration voltage.

Dependence of the fundamental and resonance frequencies on the position of the movable short L1. Open circles and closed triangles represent measured and calculated frequencies, respectively. The fundamental mode at is used as the acceleration voltage.

Comparison between the measured values and those calculated by the MAFIA code at the fundamental frequencies. The calculated values are multiplied by a factor of 0.7.

Comparison between the measured values and those calculated by the MAFIA code at the fundamental frequencies. The calculated values are multiplied by a factor of 0.7.

Dependence of the fifth-harmonic resonance frequency on the position of L5 for various dimensions of the FT resonator. The dimensions a and b represent inner- and outer-tube diameters, respectively. The gap of C5 was fixed at 25 mm during the simulation.

Dependence of the fifth-harmonic resonance frequency on the position of L5 for various dimensions of the FT resonator. The dimensions a and b represent inner- and outer-tube diameters, respectively. The gap of C5 was fixed at 25 mm during the simulation.

Dependence of position setting of the C5 gap (circles) and the L5 position (squares) on the resonance frequency for impedance matching. Open and closed symbols represent calculated and measured positions, respectively. In the calculation, optimum L5 positions were searched for fixed C5 gap of 6, 25, and 50 mm. The required resonance range was completely covered by the actual FT resonator.

Dependence of position setting of the C5 gap (circles) and the L5 position (squares) on the resonance frequency for impedance matching. Open and closed symbols represent calculated and measured positions, respectively. In the calculation, optimum L5 positions were searched for fixed C5 gap of 6, 25, and 50 mm. The required resonance range was completely covered by the actual FT resonator.

Comparison between the measured value (closed circles) and the value calculated by the MAFIA code for the fifth-harmonic frequency. The calculated values (open circles) are multiplied by a factor of 0.5.

Comparison between the measured value (closed circles) and the value calculated by the MAFIA code for the fifth-harmonic frequency. The calculated values (open circles) are multiplied by a factor of 0.5.

Variations in the measured values and the transmission level S21 for the different settings of the C5 position tuned to 75 MHz. The position of L5 was shifted 118 mm through the course of the measurement.

Variations in the measured values and the transmission level S21 for the different settings of the C5 position tuned to 75 MHz. The position of L5 was shifted 118 mm through the course of the measurement.

Fifth-harmonic voltage ratios at a radius of 915 mm to the tip of the dee electrode . The extraction radius of the cyclotron is 923 mm. The voltage ratio decreases for the higher frequency because of the short wavelength of the standing wave.

Fifth-harmonic voltage ratios at a radius of 915 mm to the tip of the dee electrode . The extraction radius of the cyclotron is 923 mm. The voltage ratio decreases for the higher frequency because of the short wavelength of the standing wave.

Distribution of the number of revolutions per 100 mm for acceleration harmonics of 1, 2, and 3. Variation in the measured fifth-harmonic voltages along the cyclotron radius is also shown.

Distribution of the number of revolutions per 100 mm for acceleration harmonics of 1, 2, and 3. Variation in the measured fifth-harmonic voltages along the cyclotron radius is also shown.

Picture of newly developed deflector probe with a tungsten sheet and schematic explanation of the radial turn pattern measurement.

Picture of newly developed deflector probe with a tungsten sheet and schematic explanation of the radial turn pattern measurement.

Radial current distribution at 220 MeV measured around the entrance of the deflector electrode. Clear turn separation was observed for FT acceleration.

Radial current distribution at 220 MeV measured around the entrance of the deflector electrode. Clear turn separation was observed for FT acceleration.

Distribution of the 260 MeV beam current measured by an integral current probe around the entrance of the magnetic channel. The extraction efficiency at the deflector electrode was estimated to be 97% for FT acceleration and 79% for the fundamental acceleration.

Distribution of the 260 MeV beam current measured by an integral current probe around the entrance of the magnetic channel. The extraction efficiency at the deflector electrode was estimated to be 97% for FT acceleration and 79% for the fundamental acceleration.

Distribution of the beam current at 45 MeV measured by the differential current probe. The extraction efficiency was improved from 56% to 86% by FT acceleration.

Distribution of the beam current at 45 MeV measured by the differential current probe. The extraction efficiency was improved from 56% to 86% by FT acceleration.

Pulse train of the beam bunch for 260 MeV at the beam transport line when a single beam bunch was injected to the cyclotron: (a) single-turn extraction with FT acceleration and (b) typical pulse train from multiturn extraction. The period of the natural beam bunches corresponds to 15 channels in the axis of time.

Pulse train of the beam bunch for 260 MeV at the beam transport line when a single beam bunch was injected to the cyclotron: (a) single-turn extraction with FT acceleration and (b) typical pulse train from multiturn extraction. The period of the natural beam bunches corresponds to 15 channels in the axis of time.

Beam current distributions of the 260 MeV measured at the focusing point of the analyzing magnet for estimation of the energy spread. The beam width was limited to 0.1 mm by a microslit system placed at the upstream position of the magnet, and the beam expansion of 0.1 mm at the downstream corresponds to 0.01% of the energy spread.

Beam current distributions of the 260 MeV measured at the focusing point of the analyzing magnet for estimation of the energy spread. The beam width was limited to 0.1 mm by a microslit system placed at the upstream position of the magnet, and the beam expansion of 0.1 mm at the downstream corresponds to 0.01% of the energy spread.

Beam current distributions for the energy spread of 0.05, 0.1, 0.3, and 0.5%, simulated by summation of Gaussian profiles with a fixed beam bunch interval and width of 3 and 1.2 mm, respectively.

Beam current distributions for the energy spread of 0.05, 0.1, 0.3, and 0.5%, simulated by summation of Gaussian profiles with a fixed beam bunch interval and width of 3 and 1.2 mm, respectively.

Article metrics loading...

Full text loading...

Commenting has been disabled for this content